The usage of mixture and mixed distributions in lifetime analysis

In this study, examined the usage of mixture and mixed distributions in survival analysis. Especially, proposed mixture distribution models instead of the standard probability distributions of inadequate modeling of heterogeneous survival. The study primarily is given the basic definitions and functions about survival analysis. Modelling survival times formed mixture models with different parameters of the same distribution, mixture models of two different distribution, and mixture model of mixed extended exponential-geometric (EEG) distributions. The generated mixture distribution models the maximum likelihood estimations of the parameters are obtained by the EM algorithm. Effectiveness parameter estimation of algorithms are determined with simulation study. In addition, have been revealed the advantages opposite standard probability distributions of the proposed mixture distribution models the data modeling using real-life data.